RANDOM FOREST CLASSIFICATION OF JAMBI AND SOUTH SUMATERA USING ALOS PALSAR DATA

Abstract

Recently, Synthetic Aperture Radar (SAR) satellite imaging has become an increasing popular data source especially for land cover mapping because its sensor can penetrate clouds, haze, and smoke which a serious problem for optical satellite sensor observations in the tropical areas. The objective of this study was to determine an alternative method for land cover classification of ALOS-PALSAR data using Random Forest (RF) classifier. RF is a combination (ensemble) of tree predictors that each tree predictor depends on the values of a random vector sampled independently and with the same distribution for all trees in the forest. In this paper, the performance of the RF classifier for land cover classification of a complex area was explored using ALOS PALSAR data (25m mosaic, dual polarization) in the area of Jambi and South Sumatra, Indonesia. Overall accuracy of this method was 88.93%, with producer’s accuracies for forest, rubber, mangrove & shrubs with trees, cropland, and water classes were greater than 92%.